Vehicle positioning method and device, storage medium and electronic device

Abstract

A vehicle positioning method and device, a computer-readable storage medium and an electronic device, wherein the method comprises: pre-positioning a target vehicle in a driving state based on a global positioning system to obtain a priori position of the target vehicle in a map (S102); obtaining driving image data collected by a camera sensor in the target vehicle in a target area where the priori position is located (S104); visually identifying the driving image data using a vector map of the target area to obtain lateral position information and longitudinal position information of the current position of the target vehicle (S106); determining a target lane where the target vehicle is located based on the lateral position information, and locating the driving position of the target vehicle in the target lane based on the longitudinal position information (S108). The method solves the technical problem of low vehicle positioning accuracy under the premise of limited cost in the prior art.

Description

[0001] This patent application claims priority to U.S. patent application No. 17 / 142,212, filed on January 5, 2021, the entirety of which is incorporated by reference into this application. Technical Field

[0002] The present invention relates to the field of vehicle control, and in particular to a vehicle positioning method and device, a storage medium and an electronic device. Background Art

[0003] Today, intelligent driving technology is being applied to an increasing number of vehicles to assist drivers in completing the driving process more safely and reliably. To achieve this, it is usually necessary to accurately locate the current position of the vehicle.

[0004] Currently, locating a vehicle's current position relies heavily on expensive onboard LiDAR (for light detection and ranging) and expensive pre-drawn high-resolution point cloud maps. That is, with the help of onboard LiDAR and high-resolution maps, the positioning system is assisted to achieve vehicle positioning. However, when using the above method, due to the high cost of using onboard LiDAR and high-resolution point cloud maps, this method cannot be widely applied to more vehicles. In other words, for most cost-constrained vehicles, accurate vehicle positioning within a controlled cost is still not possible.

[0005] In view of the above problems, no effective solution has been proposed so far. Summary of the Invention

[0006] In view of this, embodiments of the present invention provide a vehicle positioning method and apparatus, a storage medium, and an electronic device to at least solve the problem of low vehicle positioning accuracy under the premise of limited cost in the prior art.

[0007] According to one aspect of an embodiment of the present invention, a vehicle positioning method is provided, comprising: pre-positioning a target vehicle in operation based on a global positioning system to obtain a priori position of the target vehicle in a map; obtaining driving image data collected by a camera sensor in the target vehicle within a target area where the priori position is located; visually identifying the driving image data using a vector map of the target area to obtain lateral position information and longitudinal position information of the target vehicle's current position; determining a target lane where the target vehicle is located based on the lateral position information, and locating the driving position of the target vehicle within the target lane based on the longitudinal position information.

[0008] According to another aspect of an embodiment of the present invention, a vehicle positioning device is also provided, including: a pre-positioning unit, used to pre-position a target vehicle in a running state based on a global positioning system, and obtain the prior position of the target vehicle in a map; an acquisition unit, used to obtain driving image data collected by a camera sensor in the target vehicle in a target area where the prior position is located; an identification unit, used to perform visual recognition on the driving image data using a vector map of the target area, to obtain lateral position information and longitudinal position information of the current position of the target vehicle; a positioning unit, used to determine the target lane where the target vehicle is located based on the lateral position information, and to locate the driving position of the target vehicle in the target lane based on the longitudinal position information.

[0009] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, wherein the computer program is configured to execute the vehicle positioning method during operation.

[0010] According to another aspect of an embodiment of the present invention, an electronic device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to execute the vehicle positioning method through the computer program.

[0011] In an embodiment of the present invention, a camera sensor replaces an on-board laser radar (LIDAR) to capture driving image data of a target vehicle in operation. Combined with a vector map of the target vehicle's current location, this data is used to obtain lateral and longitudinal position information of the target vehicle's current location. This lateral and longitudinal position information is then used to determine the target vehicle's current lane and its driving position within the target lane. In other words, while reducing operational costs, the driving image data and vector map collected by the camera sensor are combined to accurately identify the target lane and driving position of the target vehicle in operation, thereby improving vehicle positioning accuracy. This overcomes the relatively low vehicle positioning accuracy inherent in existing technologies, which suffer from limited costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The accompanying drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. These drawings are used to explain the present invention together with the embodiments of the present invention, rather than to limit the present invention. In the accompanying drawings:

[0013] Figure 1 is a flow chart of an optional vehicle positioning method according to an embodiment of the present invention;

[0014] Figure 2 is a schematic diagram of an application environment of an optional vehicle positioning method according to an embodiment of the present invention;

[0015] Figure 3 is a schematic diagram of an optional vehicle positioning method according to an embodiment of the present invention;

[0016] Figure 4 is a schematic diagram of another optional vehicle positioning method according to an embodiment of the present invention;

[0017] Figure 5 is a schematic diagram of another optional vehicle positioning method according to an embodiment of the present invention;

[0018] Figure 6 2 is a schematic structural diagram of an optional vehicle positioning device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to enable those skilled in the art to better understand the technical solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative work should fall within the scope of protection of the present invention.

[0020] It should be noted that the terms "first", "second", etc. in the specification, claims, and drawings of the present invention are used to distinguish similar objects, but are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or are inherent to these processes, methods, products, or apparatus.

[0021] According to one aspect of an embodiment of the present invention, a vehicle positioning method is provided. Optionally, as an optional implementation, as Figure 1 As shown, the vehicle positioning method includes:

[0022] S102, pre-positioning the target vehicle in a running state based on the global positioning system to obtain a priori position of the target vehicle in the map;

[0023] S104, obtaining driving image data collected by a camera sensor in the target vehicle within the target area where the prior position is located;

[0024] S106, performing visual recognition on the driving image data using a vector diagram of the target area to obtain lateral position information and longitudinal position information of the current position of the target vehicle;

[0025] S108 , determining the target lane where the target vehicle is located according to the lateral position information, and locating the driving position of the target vehicle in the target lane according to the longitudinal position information.

[0026] Optionally, in this embodiment, the vehicle positioning method can be applied to, but not limited to, vehicles equipped with advanced driver assistance systems (ADAS) to assist the vehicle in achieving accurate positioning in real time during autonomous driving. The ADAS system uses various sensors installed on the vehicle to perceive the surrounding environment at any time while the vehicle is driving, collect data, identify, detect and track static and dynamic objects, collect map data from the navigator, and perform calculations and analysis on the system, so that the driver can be aware of possible dangers in advance, thereby effectively improving the comfort and safety of vehicle driving. The sensor can be, but is not limited to, a low-cost camera sensor, and the map data can be, but is not limited to, a vector map. Among them, size information and direction information are both recorded in the vector map. The above is just an example, and this embodiment is not limited to this.

[0027] In this embodiment, a camera sensor replaces an on-board laser radar (LIDAR) to capture driving image data of a target vehicle while in motion. Combined with a vector map of the target vehicle's current location, this data is used to obtain lateral and longitudinal position information of the target vehicle's current position. This lateral and longitudinal position information is then used to determine the target vehicle's current lane and its driving position within the target lane. In other words, while reducing costs, the driving image data collected by the camera sensor and the vector map are combined to accurately identify the target vehicle's current lane and driving position, thereby improving vehicle positioning accuracy. This overcomes the relatively low vehicle positioning accuracy inherent in existing technologies, which often face cost constraints.

[0028] Optionally, in this embodiment, the global positioning system may include a global navigation satellite system (GNSS), and the vector map may be obtained by the following components: an inertial measurement unit (IMU) and an odometer. GNSS is a space-based radio navigation system that can provide users with all-weather three-dimensional coordinates, velocity, and time information at any location on or near the Earth's surface. An IMU typically consists of a gyroscope, an accelerometer, and an algorithm processing unit, and obtains a human body's motion trajectory by measuring acceleration and rotation angle.

[0029] In addition, in this embodiment, in the process of visually identifying the driving image data using the vector map of the area where the target vehicle is currently located to obtain the lateral position information and longitudinal position information of the target vehicle's current location, the method may include but is not limited to the use of visual recognition technology and deep learning implemented by a deep neural network.

[0030] The visual recognition technology in this embodiment is a machine vision technology that mainly uses computers to simulate human visual functions, extract information from target images, and process and understand them, ultimately for actual detection, measurement, and control.

[0031] The deep neural network in this embodiment simulates the human brain's thinking model and, in combination with a vector map, identifies lane information for each lane on the road, as determined by the deep neural network, from the target vehicle's driving image data. The lane information herein may include, but is not limited to: the lane markings for each lane, the lane marking of the target lane currently in which the target vehicle is located, the offset distance of the target vehicle relative to the lane center of the target lane; and also includes the target vehicle's heading angle and the error between the azimuth angle and the tangent of the target lane. The above lane information is merely an example; this embodiment may further include other information, which is not limited here.

[0032] According to an embodiment of the present invention, a camera sensor replaces an on-board laser radar to acquire driving image data of a target vehicle in operation. Combined with a vector map of the target vehicle's current location, the lateral and longitudinal position information of the target vehicle's current location is obtained. This lateral and longitudinal position information, combined with the driving image data and vector map collected by the camera sensor, accurately identifies and locates the target lane and driving position of the target vehicle in operation, thereby improving vehicle positioning accuracy. This overcomes the relatively low vehicle positioning accuracy inherent in existing technologies, which often suffers from limited cost.

[0033] As an optional solution, the steps of performing visual recognition on driving image data using a vector map of the target vehicle's current location to obtain lateral position information and longitudinal position information of the target vehicle's current location include:

[0034] S1, using visual recognition technology to perform visual recognition on the driving image data to obtain a recognition result;

[0035] S2, matching the vector map with the recognition result to determine road information of the area where the target vehicle is currently located in the driving image data, wherein the road information includes a road network and fixed road objects located beside the road;

[0036] S3, obtaining the lateral position information and longitudinal position information of the current position of the target vehicle according to the road information.

[0037] Optionally, in this embodiment, the road network may include, but is not limited to, a road network formed by various lanes in different directions within an area. The aforementioned fixed road objects may include, but are not limited to, traffic lights, traffic signs, utility poles, etc. The above are merely examples and are not intended to limit this embodiment.

[0038] Specifically, refer to Figure 2 In the scenario shown, it is assumed that the recognition results are obtained by using visual recognition technology on driving image data collected by the camera sensor. After comparing the recognition results with the information in the ADAS vector map, road information such as the road network and fixed road objects (such as traffic lights and signs) in the area where the target vehicle 202 is located is obtained. In addition, this road information is used to determine the lateral and longitudinal position information of the target vehicle 202.

[0039] Among them, in an accurate vector map, such as OpenDrive, various key geometric information related to a road segment (such as the current lane turning radius, current lane marking, etc.) can be intuitively obtained by retrieving or zooming from a given reference point on the map. Figure 2 As shown, the elliptical 2D Gaussian distribution combines the traditional global positioning system for inertial measurement unit and odometry measurement, which is demonstrated in the figure by the Gaussian kernel size. Therefore, based on the above comprehensive information, the lane width can be identified as larger than 3.5m through analysis and estimation. In addition, as Figure 2 As shown, it can also be determined that the target lane currently located by the target vehicle 202 is the outermost lane (ie, lateral position information), and the driving position is a curve (ie, longitudinal position information).

[0040] The embodiment of the present application combines visual recognition results and vector graphics to accurately locate the lateral position information and longitudinal position information of the target vehicle, thereby ensuring improved positioning accuracy while saving usage costs.

[0041] As an optional solution, the step of obtaining the lateral position information of the current position of the target vehicle according to the road information includes:

[0042] S1, using a deep neural network to detect and identify road information to obtain lane information of all lanes on the road where the target vehicle is currently traveling, wherein the lane information includes: lane lines marking each lane, a lane marker of the target lane where the target vehicle is currently located, and an offset distance of the target vehicle relative to the lane center of the current target lane;

[0043] S2, generating lateral position information according to the lane mark and offset distance of the target lane.

[0044] Optionally, in this embodiment, after obtaining lane information of all lanes on the road on which the target vehicle is currently traveling, the method further includes: using an independent coding channel or bit mask to differentially mark each lane on the road on which the target vehicle is currently traveling in a different marking manner.

[0045] It should be noted that the deep neural network described here is trained using multiple sample data sets and is used to identify and compare lane markings across multiple lanes in road information. The network structure of the deep neural network described herein may include, but is not limited to, convolutional layers, pooling layers, and fully connected layers. Each weight parameter in the described network structure can be obtained through multiple iterations of training. This description will not be repeated here.

[0046] In addition, in this embodiment, lane lines of multiple lanes (e.g., straight lanes or curved lanes) within the camera field of view are obtained by detecting and identifying road information. The lane lines here can be marked in different ways, but are not limited thereto.

[0047] For example, assuming that the lane information of all lanes on the driving road where the target vehicle 302 is located can be as follows Figure 3 As shown: Figure 3 As shown in (a), each lane line can be represented by a dot. Figure 3 As shown in (b), different lane lines can also be identified by using different line types. In addition, different lane lines can also be marked with different colors (not shown in the figure). To this end, in this embodiment, it is possible but not limited to using additional encoding channels or separate bit masks (consistent with the size of the captured image) to implement a time-consuming pattern extraction process that eliminates the process associated with each individual lane line, for example, it is not necessary to slide the window again to identify each pixel in the image.

[0048] In addition, if Figure 3 As shown, the identified lane information may also include street lights 304 , traffic lights 306 , and the like. Figure 3 An example is shown, and it is not limited to this embodiment.

[0049] Please refer to the following examples for detailed description:

[0050] After visually recognizing the driving image data to obtain a recognition result and matching the recognition result with the vector map to obtain road information, a deep neural network can be used to detect and identify the road information to obtain lane information for each lane. The algorithm module in the vehicle's onboard processor fits the curve of each lane line, and then generates lane parameters for each curve based on a fitting function (such as a quadratic polynomial curve fitting function, a cubic polynomial curve fitting function, or cubic spline interpolation). The lane parameters here may include but are not limited to lane markings and offset distances from the lane center.

[0051] Furthermore, visual recognition technology is used to determine the lane marker of the target lane in which the target vehicle is currently located, as well as the target vehicle's offset distance from the lane center of the target lane, from all lanes. For example, assuming the target vehicle's target lane is identified as an exit, inner lane, or outer lane, and is located on the inside of the lane, the target vehicle is located at a position offset by 10 centimeters from the lane center of the target lane.

[0052] Through the embodiments of the present application, a deep neural network is used to detect and identify the road, obtain lane information for each lane, and then identify the lateral position information of the target lane where the target vehicle is located. For example, the lane marker of the target lane and the offset distance of the target vehicle relative to the lane center of the target lane can be accurately located laterally, with centimeter-level accuracy.

[0053] As an optional solution, obtaining the longitudinal position information of the current position of the target vehicle according to the road information includes:

[0054] S1, when acquiring the lateral position information, acquiring lane parameters of a fitted lane corresponding to a target lane where the target vehicle is located according to the road information, wherein the lane parameters include a radius of the target lane;

[0055] S2, reads the reference lane radius of the target lane from the vector map;

[0056] S3, compare the target lane radius with the reference lane radius;

[0057] S4, determining the driving position of the target vehicle in the target lane according to the comparison result to obtain longitudinal position information of the target vehicle.

[0058] It is described specifically with reference to the following embodiments: When the target lane in which the target vehicle is located is determined, fitting parameters for fitting the lane can be obtained, wherein the lane parameters used for fitting the lane here may include but are not limited to: a fitting coefficient, an offset distance from the center of the lane, an error between the heading angle of the target vehicle and the tangent direction of the lane, and a turning radius of the current position of the target vehicle.

[0059] Assume that the lateral positioning result of the target vehicle is as follows Figure 4 As shown, the target vehicle is located in the right outer lane, and the lane radius here is radius R (its positioning accuracy is in the order of centimeters). Figure 4 The narrow boundary of the lateral Gaussian distribution (the ellipse where the target vehicle is located).

[0060] In addition, the vector diagram retrieves the reference lane radii r1, r2, ..., rn, recorded at multiple consecutive locations in the longitudinal direction of the target lane based on the target vehicle's current position. By sequentially comparing each radius R with the reference lane radii r1, r2, ..., rn, a closest-matching mechanism in the longitudinal direction is employed to accurately locate the target vehicle's longitudinal position in the current target lane. Specifically, the longitudinal position corresponding to the reference lane radius ri closest to lane radius R is determined as the target vehicle's driving position.

[0061] The embodiment of the present application determines the corresponding driving position of the target vehicle in the target lane by comparing the lane radius, thereby achieving accurate positioning of the target vehicle in the longitudinal direction.

[0062] As an optional solution, the driving position of the target vehicle in the target lane is determined based on the comparison result to obtain the longitudinal position information of the target vehicle, including:

[0063] S1, determining the longitudinal driving area of ​​the target vehicle in the target lane according to the comparison result;

[0064] S2, determining a driving position according to the longitudinal driving area.

[0065] Optionally, in this embodiment, the solution for determining the driving position according to the longitudinal driving area may include at least one of the following solutions:

[0066] 1. In the case where the lane curvature of the target lane is not constant, a maximum likelihood method is used to determine the driving position based on the longitudinal driving area using the following data: the actual curvature of the target lane, the reference curvature estimated for the target lane by the positioning system, and the sampled curvature estimated by the target vehicle based on the sample points associated with the current position;

[0067] 2. If the longitudinal travel area includes map landmarks, the travel position is estimated based on the target landmark closest to the current position of the target vehicle;

[0068] 3. Estimating the driving position using the positioning marker data stored in the positioning system, provided that the lane curvature of the target lane is constant and that the longitudinal driving area does not include map landmarks;

[0069] 4. Estimate the driving position from the longitudinal driving area using carrier phase lookup technology.

[0070] It should be noted that, in the above embodiment, since the processing cost is successively higher than that of the lane radius method, the longitudinal travel area of ​​the target vehicle can be first determined by using the above method, and further, the longitudinal travel position of the target vehicle can be estimated based on the longitudinal travel area by using any of the following methods:

[0071] 1) In road sections with non-constant (irregular) curvature, the maximum likelihood method can be used to solve the longitudinal positioning problem. Figure 5 As shown, curve 502 represents the actual road with true curvature on the map; Gaussian distribution curve 504 represents the longitudinal position estimated based on the fusion of the global navigation satellite system, inertial measurement unit, and odometer. As can be seen from the figure, there is a certain deviation between the two, for example, approximately 10.0 meters. The diamond points in curve 506 represent the curvature estimated from a set of sampling points around the current position of the target vehicle. Using these curves 502 to 506, the maximum likelihood method can be used to estimate the precise longitudinal position within 1-2 meters.

[0072] 2) Within a road section marked with mapped landmarks, the system uses visual recognition technology to identify the landmarks (such as utility poles) closest to the target vehicle's location. These recognition results are then used to update its longitudinal estimate. Based on visual performance, camera viewing angle, processing speed, and map accuracy, the system achieves precise longitudinal positioning within 2-3 meters.

[0073] 3) Combining the above-mentioned Examples 1)-2) for co-localization.

[0074] 4) On a road section with constant curvature (for example, a straight line or a circle), the above-mentioned method 2 can be used to directly perform longitudinal positioning.

[0075] 5) On road sections with constant curvature and no obvious road signs, other visual cues can be obtained to improve positioning of GNSS, IMU, odometer, etc. For example, city / skyline sketches can be used as long as they are represented in some way on the map.

[0076] 6) Only when none of the above tools are available, GNSS, inertial measurement units, and odometers are used. Furthermore, real-time kinematic (RTK) technology is used in parallel with all of the above mechanisms. Other radio-based positioning enhancements, such as V2X mechanisms, can also be used to achieve the target vehicle's own positioning.

[0077] The embodiment of the present application further performs precise longitudinal positioning of the longitudinal driving area by combining different methods, thereby obtaining the driving position of the target vehicle in the target lane, thereby ensuring the accuracy of vehicle positioning.

[0078] It should be noted that, for the sake of simplicity, the aforementioned method embodiments are described as a series of action combinations. However, those skilled in the art should be aware that the present invention is not limited by the order of the actions described, because according to the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in this specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0079] According to another aspect of the present invention, a vehicle positioning device for implementing the above vehicle positioning method is also provided. Figure 6 As shown, the device includes:

[0080] 1) A pre-positioning unit 602 pre-positions a target vehicle in a running state based on a global positioning system to obtain a priori position of the target vehicle in a map;

[0081] 2) an acquisition unit 604 for acquiring driving image data collected by a camera sensor in a target vehicle within a target area where the prior position is located;

[0082] 3) an identification unit 606 for performing visual recognition on the driving image data using a vector map of the target area to obtain lateral position information and longitudinal position information of the target vehicle's current location;

[0083] 4) A positioning unit 608 is used to determine the target lane where the target vehicle is located based on the lateral position information, and to locate the driving position of the target vehicle in the target lane based on the longitudinal position information.

[0084] Optionally, for the specific embodiment of the vehicle positioning device provided in the embodiment of the present application, reference may be made to the above method embodiment, which will not be repeated here.

[0085] As an optional solution, the identification unit includes:

[0086] 1) Recognition module, used to perform visual recognition on driving image data using visual recognition technology to obtain recognition results;

[0087] 2) a matching module, configured to match the vector map with the recognition result to determine road information of the area where the target vehicle is currently located in the driving image data, wherein the road information includes the road network and fixed road objects located beside the road;

[0088] 3) An acquisition module is used to acquire the lateral position information and longitudinal position information of the current position of the target vehicle based on the road information.

[0089] Optionally, for the specific embodiment of the vehicle positioning device provided in the embodiment of the present application, reference may be made to the above method embodiment, which will not be repeated here.

[0090] According to one aspect of the present application, a computer program product or computer program is provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the above-described vehicle positioning method, wherein the computer program is configured to perform the steps of any of the above-described method embodiments when executed.

[0091] Optionally, in this embodiment, the computer-readable storage medium may be arranged to store a computer program for performing the following steps:

[0092] S1, pre-positioning the target vehicle in the running state based on the global positioning system to obtain the prior position of the target vehicle in the map;

[0093] S2, obtaining driving image data collected by a camera sensor in the target vehicle within the target area where the prior position is located;

[0094] S3, performing visual recognition on the driving image data using the vector map of the target area to obtain the lateral position information and the longitudinal position information of the current position of the target vehicle;

[0095] S4, determining the target lane where the target vehicle is located according to the lateral position information, and locating the driving position of the target vehicle in the target lane according to the longitudinal position information.

[0096] Alternatively, in this embodiment, those skilled in the art will appreciate that all or part of the steps of the method in the above embodiment may be performed by a program instructing hardware related to the terminal device. The program may be stored in a computer-readable storage medium, and the storage medium may include a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0097] The serial numbers of the embodiments of the present invention are only used for description and do not represent the advantages or disadvantages of the embodiments.

[0098] If the integrated units in the above embodiments are implemented in the form of software functional units and sold or used as independent products, they can be stored in the above-mentioned computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the existing technology, or all or part of the technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes a number of instructions for causing one or more computer devices (such as personal computers, servers, or network devices) to execute all or part of the steps of the methods described in various embodiments of the present invention.

[0099] In the above embodiments of the present invention, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0100] In the several embodiments provided in this application, it should be understood that the disclosed client can be implemented in other ways. Among them, the device embodiments described above are merely exemplary. For example, the division of the units is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be achieved through some interfaces, and the indirect coupling or communication connection between units or modules can be achieved in electrical or other forms.

[0101] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present invention.

[0102] In addition, the functional units in the embodiments of the present invention may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0103] The above are only preferred embodiments of the present invention. It should be noted that, for those skilled in the art, further improvements and modifications can be made without departing from the principles of the present invention, and such improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A vehicle positioning method, comprising: Pre-positioning a target vehicle in operation based on a global positioning system to obtain a priori position of the target vehicle in a map; Acquire driving image data collected by a camera sensor in the target vehicle within the target area where the priori position is located; Performing visual recognition on the driving image data using visual recognition technology to obtain a recognition result; Matching the vector map of the target area with the recognition result to determine road information of the area where the target vehicle is currently located in the driving image data, wherein the road information includes a road network and fixed road objects located beside the road; Acquire lateral position information of the current position of the target vehicle according to the road information; When acquiring the lateral position information, acquiring lane parameters of a fitting lane corresponding to a target lane where the target vehicle is located according to the road information, wherein the lane parameters include a target lane radius; Reading a reference lane radius of the target lane from the vector diagram; comparing the target lane radius with the reference lane radius; Determine the driving position of the target vehicle in the target lane according to the comparison result to obtain the longitudinal position information of the target vehicle; The target lane where the target vehicle is located is determined according to the lateral position information, and the driving position of the target vehicle in the target lane is located according to the longitudinal position information.

2. The method according to claim 1, wherein the step of acquiring the lateral position information of the current position of the target vehicle according to the road information comprises: Detecting and identifying the road information using a deep neural network to obtain lane information of all lanes on the road on which the target vehicle is currently traveling, wherein the lane information includes: lane lines marked for each lane, a lane marker of the target lane where the target vehicle is currently located, and an offset distance of the target vehicle relative to the lane center of the current target lane; The lateral position information is generated according to the lane identifier and the offset distance of the target lane.

3. The method according to claim 1, wherein determining the driving position of the target vehicle in the target lane according to the comparison result to obtain the longitudinal position information of the target vehicle comprises: Determining a longitudinal driving area of ​​the target vehicle in the target lane according to the comparison result; The driving position is determined based on the longitudinal driving range.

4. The method according to claim 3, wherein determining the driving position according to the longitudinal driving area comprises at least one of the following: In the case where the lane curvature of the target lane is not constant, using a maximum likelihood method on the following data to determine the driving position according to the longitudinal driving area: an actual curvature of the target lane, a reference curvature estimated for the target lane by a positioning system, and a sampled curvature estimated by the target vehicle based on a sampled point associated with a current position; In the case where the longitudinal travel area includes map landmarks, estimating the travel position based on a target landmark closest to the current position of the target vehicle; estimating the driving position using positioning marker data stored in the positioning system when the lane curvature of the target lane is constant and the longitudinal driving area does not include map landmarks; The travel position is estimated from the longitudinal travel region using a carrier phase lookup technique.

5. The method according to claim 2, wherein after acquiring lane information of all lanes on the road on which the target vehicle is currently traveling, the method further comprises: Each lane on the road on which the target vehicle is currently traveling is differentially marked in a different marking manner using independent coding channels or bit masks.

6. A vehicle positioning device, comprising: A pre-positioning unit, used to pre-position a target vehicle in a running state based on a global positioning system, and obtain a priori position of the target vehicle in a map; An acquisition unit, used for acquiring driving image data collected by a camera sensor in the target vehicle in the target area where the priori position is located; a recognition unit, configured to perform visual recognition on the driving image data using visual recognition technology to obtain a recognition result; match the vector map of the target area with the recognition result to determine road information of the area where the target vehicle is currently located in the driving image data, wherein the road information includes a road network and fixed road objects located beside the road; obtain lateral position information of the current position of the target vehicle according to the road information; when obtaining the lateral position information, obtain lane parameters of a fitting lane corresponding to a target lane where the target vehicle is located according to the road information, wherein the lane parameters include a radius of the target lane; Reading a reference lane radius of the target lane from the vector diagram; comparing the target lane radius with the reference lane radius; Determine the driving position of the target vehicle in the target lane according to the comparison result to obtain the longitudinal position information of the target vehicle; A positioning unit is used to determine the target lane where the target vehicle is located according to the lateral position information, and to locate the driving position of the target vehicle in the target lane according to the longitudinal position information.

7. A computer-readable storage medium, wherein the computer-readable storage medium comprises a stored program, wherein: The method according to claim 1 is executed when the program is executed.

Images